Skip to main content

FUJITSU JOURNAL

Search

Next-generation Internet: LOD Represents the Future of Data

Technology transforming the Internet into a database of knowledge

The emergence of the Internet has had a profound impact on our lives as it connects all people through a single network, the first time this has been possible in the history of humanity.
Today, a new revolutionary technology with an impact comparable to that of the Internet is attracting attention globally, Linked Open Data (LOD).
LOD refers to a system for linking everything by opening up data to the public in formats that can be used by everyone. This will develop the Internet into a database of knowledge completely different from previous networks.

After LOD will come artificial intelligence, which is capable of effectively using website data such as text, images, and sounds to respond to surrounding situations and human emotion.
In the future, a world resembling science fiction may arrive in which artificially intelligent robots support us in our daily lives.

As information becomes LOD, there will be great potential to realize this type of artificial intelligence.

Research Interview

LOD utilizes the entire Web as an enormous database

As exemplified by the Open Data Charter signed by G8 leaders, open data initiatives are underway. In line with these initiatives, there are already widespread movements to make information held by various parties, from national and local governments to academic groups, available to the public.

LOD is an open data network by which various kinds of information are opened to the public in data formats that make it easy for computers to process and link data items to each other; the network can be used by anyone for secondary purposes. This technology enables users to treat the entire Internet as a single database and perform much more advanced searches than is possible on websites today.

At present, websites consist of data prepared to facilitate human comprehension such as text, CSV files, and PDF files, however it is difficult to process these types of data mechanically.

For example, in order to identify the company with the highest profit ratio per employee in the world, it is necessary to perform several searches using keywords such as number of employees, profit ratios, and top ranking globally and then to compare the gathered data. In addition, since these searches will be performed by people, it is difficult to confirm all information on the relevant websites, and there is a possibility that some important information may be overlooked.

By contrast, LOD enables machine batch processing of all types of data because all related data is multiply linked to other data to form a single data structure.

With LOD, by searching using a single query, "the company with the highest profit ratio per employee in the world," the correct result will be produced out of a data set put together by extracting only numbers of employees and profit ratios.

Currently, a British public broadcasting station that manages data on more than 1,000 programs per day is using LOD as an online platform to enable it to easily discover where information about programs it wants to know is located.

What Fujitsu's LOD utilization platform can do

In order to use LOD effectively, development of superior search technologies is indispensable. So far, however, full use of LOD has not yet been made effectively because not all open data in the world has been linked. High data processing capacity is also essential because LOD requires complicated processing of combinations of vast amounts of data. Under the assumption that LOD will be used in various ways in the real world, Fujitsu is focusing its efforts on developing LOD technologies and methods to use LOD ahead of other companies.

Case studies

Case 1Analysis of businesses using LOD effectively (financial services)

Fujitsu's LOD utilization platform enables users to instantaneously obtain various kinds of data related to information on businesses around the world, from basic information; industry type and number of employees, to financial data; sales, profits, and stock prices, even media; newspaper articles. This platform facilitates multifaceted analysis of corporate performance without omitting detailed information. Furthermore, through data standardization, it simplifies routine analytical work and responds swiftly to needs for updating analytical data and comparing differences, thus bringing about innovation in business.

Case 2Promotion of regional economies through LOD (local governments)

Fujitsu's LOD utilization platform helps local governments and residents to develop tourist information into LOD in order to provide tourist information services. This contributes mainly to increasing the number of visitors, using tourist resources effectively, and revitalizing local industries by not only promoting famous tourist spots but also identifying hidden attractions in local communities off the beaten track.

LOD search technology independently developed by Fujitsu

Development of technology to allow high-speed processing of huge volumes of complicated data is indispensable for effective use of LOD. To do so, Fujitsu has conducted joint research with Insight Centre (formerly the Digital Enterprise Research Institute), an Irish research institute that is a leader in LOD research. The two organizations jointly developed a new platform to achieve high-speed, simultaneous data processing and were the first companies in the world to open their results to the public.

Although LOD is being prepared at various locations around the world, one current problem is that each data set is not linked to others, preventing users from identifying websites where they can find the data they want. Some websites do not offer search functions, making it difficult to search all LOD.

To solve this problem, through its joint research with the Insight Centre, Fujitsu successfully consolidated LOD from around the world (tens of billions of items) into a single location, thus establishing a system that enables users to simultaneously search for the information they need without visiting the relevant individual websites.

At the same time, Fujitsu succeeded in gaining search results five to ten times as fast as previous methods due to efforts in addressing problems in search structures, interlinked data takes significant time because it requires complicated comparison processing.

Back to top